Thermal Atomic Layer Etching of SiO2 by a âConversion-Etch

Subscriber access provided by University of Newcastle, Australia

Article 2

Thermal Atomic Layer Etching of SiO by a “Conversion-Etch” Mechanism Using Sequential Reactions of Trimethylaluminum and Hydrogen Fluoride Jaime W. DuMont, Amy Marquardt, Austin Cano, and Steven M. George ACS Appl. Mater. Interfaces, Just Accepted Manuscript • DOI: 10.1021/acsami.7b01259 • Publication Date (Web): 27 Feb 2017 Downloaded from http://pubs.acs.org on March 9, 2017

Just Accepted “Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted online prior to technical editing, formatting for publication and author proofing. The American Chemical Society provides “Just Accepted” as a free service to the research community to expedite the dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been fully peer reviewed, but should not be considered the official version of record. They are accessible to all readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors or consequences arising from the use of information contained in these “Just Accepted” manuscripts.

ACS Applied Materials & Interfaces is published by the American Chemical Society. 1155 Sixteenth Street N.W., Washington, DC 20036 Published by American Chemical Society. Copyright © American Chemical Society. However, no copyright claim is made to original U.S. Government works, or works produced by employees of any Commonwealth realm Crown government in the course of their duties.

Page 1 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

ACS Applied Materials & Interfaces

Revised ACS Applied Materials & Interfaces Manuscript ID am-2017-012594 February 25, 2017

Thermal Atomic Layer Etching of SiO2 by a “Conversion-Etch” Mechanism Using Sequential Reactions of Trimethylaluminum and Hydrogen Fluoride Jaime W. DuMont1, Amy E. Marquardt1, Austin M. Cano1 and Steven M. George1,2 1

Department of Chemistry and Biochemistry, 2Department of Mechanical Engineering, University of Colorado at Boulder, Colorado 80309

ABSTRACT: The thermal atomic layer etching (ALE) of SiO2 was performed using sequential reactions of trimethylaluminum (TMA) and hydrogen fluoride (HF) at 300°C. Ex situ x-ray reflectivity (XRR) measurements revealed that the etch rate during SiO2 ALE was dependent on reactant pressure. SiO2 etch rates of 0.027, 0.15, 0.20, and 0.31 Å/cycle were observed at static reactant pressures of 0.1, 0.5, 1.0 and 4.0 Torr, respectively. Ex situ spectroscopic ellipsometry (SE) measurements were in agreement with these etch rates versus reactant pressure. In situ Fourier transform infrared (FTIR) spectroscopy investigations also observed SiO2 etching that was dependent on the static reactant pressures. The FTIR studies showed that the TMA and HF reactions displayed self-limiting behavior at the various reactant pressures. In addition, the FTIR spectra revealed that an Al2O3/aluminosilicate intermediate was present after the TMA exposures. The Al2O3/aluminosilicate intermediate is consistent with a “conversion-etch” mechanism where SiO2 is converted by TMA to Al2O3, aluminosilicates and reduced silicon species following a family of reactions represented by 3SiO2 + 4Al(CH3)3 → 2Al2O3 + 3Si(CH3)4. Ex situ x-ray photoelectron spectroscopy (XPS) studies confirmed the reduction of silicon species after TMA exposures. Following the conversion reactions, HF can fluorinate the Al2O3 and aluminosilicates to species such as AlF3 and SiOxFy. Subsequently, TMA can remove 1 ACS Paragon Plus Environment


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the AlF3 and SiOxFy species by ligand-exchange transmetalation reactions and then convert additional SiO2 to Al2O3. The pressure-dependent conversion reaction of SiO2 to Al2O3 and aluminosilicates by TMA is critical for thermal SiO2 ALE. The “conversion-etch” mechanism may also provide pathways for additional materials to be etched using thermal ALE.

KEYWORDS: atomic layer etching, SiO2, trimethylaluminum, hydrogen fluoride, Al2O3, aluminosilicate.

I. INTRODUCTION Silicon dioxide (SiO2) is the most abundant mineral in the Earth’s crust. SiO2 is also an important semiconductor material. SiO2 has many uses, such as for a gate oxide, interlayer dielectric or spacer, in semiconductor devices. SiO2 etching is needed for multiple steps during semiconductor manufacturing.1 The progressive miniaturization of semiconductor devices to sub-10 nm dimensions has imposed more demanding requirements on etching. As device structures become increasingly smaller and more complex, high precision etching is necessary at the sub-1 nm level using atomic layer etching (ALE) methods.2 There are many approaches for SiO2 etching using HF/H2O solutions, HF/H2O vapor mixtures and halogen plasmas. Etching SiO2 in aqueous HF solutions is one of the oldest and useful SiO2 etching methods.3 The overall chemical reaction during SiO2 etching in HF solutions is SiO2 + 6HF → H2SiF6 + 2H2O.4-5 The mechanism of SiO2 etching in HF solutions involves the progressive fluorine replacement of hydroxyl groups on Si-OH silanol surface species using HF or HF2- to form Si-F surface species.4-6 H2O or HF can also break the Si-O-Si siloxane linkage in SiO2 to produce additional Si-OH silanol and Si-F surface species.

2 ACS Paragon Plus Environment

Page 2 of 49

Page 3 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


SiO2 can also be etched by HF/H2O in the gas phase. The gas phase approach is an extension of the solution phase approach and occurs when H2O and HF are present at pressures sufficient to form a condensed HF/H2O layer on the SiO2 surface.7-8 H2O is essential for HF etching of SiO2.9 Vapor phase SiO2 etch rates are in agreement with the etch rates expected from the composition of the condensed HF/H2O layer present at the given HF and H2O partial pressures.7 The chemistry of vapor phase SiO2 etching is the same as observed in aqueous HF solutions. In addition, H2SiF6 is removed by the reaction H2SiF6 → SiF4 + 2HF.7 Fluorocarbon-based plasma etching can also be employed for SiO2 etching.10-12 F atoms in the plasma have been identified as the active etching reactant. SiF4 has been detected as the main Si-containing reaction product. Fluorocarbon plasmas are also used in newly developed plasma ALE methods based on an Ar plasma producing low energy Ar+ ions together with a periodic fluorocarbon-based plasma exposure.13-14 The fluorocarbon-based plasma deposits a fluorocarbon layer on the SiO2 surface. The Ar+ ion bombardment then desorbs the fluorocarbon layer and also removes a controlled amount of the SiO2 layer. Thermal ALE processes have been introduced recently for Al2O3, HfO2, ZrO2, AlF3 and AlN ALE based on sequential fluorination and ligand-exchange reactions.15-22 In these thermal reactions, the metal oxide is first fluorinated to produce the metal fluoride.17, 19, 22 The metal fluoride is then removed by a metal precursor in a ligand-exchange transmetalation reaction.17, 19, 22

In this ligand-exchange reaction, the metal precursor accepts fluorine from the metal fluoride

and donates one of its ligands to the metal fluoride. The resulting reaction product formed from the metal fluoride can then desorb from the surface if this metal species is stable and volatile.20, 22 In this paper, a new gas-phase thermal SiO2 ALE process is reported based on sequential exposures of Al(CH3)3 (Trimethylaluminum (TMA)) and HF. SiO2 film thicknesses were



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

measured versus number of ALE cycles using ex situ spectroscopic ellipsometry to obtain the SiO2 etching rates. These studies determined the linearity of the SiO2 etching and the dependence of the SiO2 etching on the TMA and HF reactant pressures. Ex situ x-ray photoelectron spectroscopy (XPS) measurements were also employed to determine the silicon oxidation state after the TMA and HF reactions. In addition, in situ Fourier transform infrared (FTIR) spectroscopy measurements were used to measure the loss of absorbance from Si-O stretching vibrations and to monitor the surfaces species present after each TMA and HF exposure at various pressures. Recent studies of ZnO ALE using TMA and HF as the reactants have revealed a new “conversion-etch” reaction mechanism where a ZnO surface layer is converted to an Al2O3 surface layer and volatile Zn(CH3)2 reaction products during the TMA exposure.23 In this investigation, the FTIR and XPS studies determined that there was a similar conversion of SiO2 to Al2O3 by TMA. The observation of Al2O3, aluminosilicate species and silicon in reduced oxidation states established that the “conversion-etch” mechanism also occurs for SiO2 ALE. In addition, the reactant pressures were critical for SiO2 ALE and larger etch rates occurred at higher reactant pressures. The generality of the “conversion-etch” mechanism could expand the number of materials that can be etched using thermal ALE methods.

II. EXPERIMENTAL SECTION 1. Static Reactor, Silicon Wafers and Ex Situ Film Analysis. ALE reactions were performed in a hot wall ALD reactor.24 A proportional-integral-derivative (PID) temperature controller (2604, Eurotherm) maintained the reaction temperature at 300°C. A capacitance manometer (Baratron 121A, MKS) monitored the pressure in the reactor. The starting SiO2


Page 4 of 49

Page 5 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


substrate was a thermal SiO2 layer with a thickness of 40 nm on Si(100) wafers with a diameter of 100 mm (University Wafer). The silicon wafers were cut into samples with dimensions of 2.5 cm x 2.5 cm before the SiO2 ALE. The SiO2 ALE reactions were accomplished using sequential static exposures of TMA (97 %, Sigma-Aldrich) and HF derived from HF-pyridine (70 wt % HF, Sigma-Aldrich). HFpyridine was transferred to a gold-plated stainless steel bubbler using a dry N2-filled glovebag. The TMA and HF-pyridine precursors were held at room temperature. Each reactant exposure consisted of a static dose of 20 s followed by a purge of 120 s with 550 sccm of ultra-high purity (UHP) N2. The total N2 gas flow of 550 sccm produced a base pressure of ~2 Torr in the reactor during the purge steps. The N2, precursor, and reaction byproducts were pumped by a mechanical pump (Pascal 2010SD, Alcatel Adixen). The ex situ X-ray reflectivity (XRR) scans were measured by a high resolution X-ray diffractometer (Bede D1, Jordan Valley Semiconductors) using Cu Kα (λ = 1.540 Å) radiation. The filament voltage and current in the X-ray tube were 40 kV and 35 mA, respectively. A 5 arcsec step size and a 10 s acquisition time were used for recording all XRR scans with a range of 300 to 4500 arcsec. The analysis software (Bede REFS, Jordan Valley Semiconductors) determined film thickness, film density and surface roughness by fitting the XRR scans. Ellipsometry was performed using a spectroscopic ellipsometer (J. A. Woollam M2000D). This ellipsometer has a spectral range of 240 to 1700 nm and utilizes an incidence angle of 75°. A Cauchy model was used to determine the film thickness. This modeling was performed using analysis software (CompleteEASE, J. A. Woollam). Thermal SiO2 layers with a thickness of 40 nm on Si(100) wafers (University Wafer) were used for XPS analysis to determine the surface species after TMA-only and TMA/HF



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

exposures. These SiO2 samples were exposed to 8 Torr of TMA for 5 minutes or 8 Torr of TMA for 5 minutes followed by 4 Torr of HF for 10 minutes. After the static reactant exposures, the samples were immediately transferred to a PHI 5600 x-ray photoelectron spectroscopy (XPS) to analyze the films. A SiO2 sample without TMA or HF exposures was also analyzed with XPS. A monochromatic Al Kα x-ray source (1486.6 eV) was used to collect survey scans with a pass energy of 93.9 eV and step size of 0.400 eV. An XPS software package (Auger Scan, RBD Instruments) was used to collect the XPS spectra. Casa XPS software (Casa XPS, Casa Software) was used for peak analysis using the C 1s, O 1s, Si 2p, Al 2p and F 1s XPS signals. All peaks were calibrated to the C 1s peak centered at 284.8 eV. The oxidation state of Si was determined through fitting the Si 2p peak profile using Gaussian-Lorentzian line shapes after a Shirley background subtraction. XPS analysis of the SiO2 control sample was used to determine the Si 2p binding energy position and full width at half maximum (FWHM) of the 4+ oxidation state peak. The binding energy (104.3 eV) and FWHM (2.1 eV) of the 4+ oxidation state were imposed in the fittings of the TMA-only and TMA/HF exposed samples. The binding energy positions for the 3+, 2+, 1+ and 0 oxidation states were defined as -1, -2, -3, and -4 eV with respect to the 4+ peak.25-26 The FWHM values of the 3+, 2+, 1+ and 0 oxidation states were set to 1.2 ± 0.1, 1.1 ± 0.1, 1.0 ± 0.1 and 1.0 ± 0.1 eV, respectively.25-27 2. In Situ Fourier Transform Infrared (FTIR) Spectroscopy Studies. The surface chemistry during sequential exposures of TMA (97 %, Sigma-Aldrich) and HF derived from HFPyridine (70 wt% HF, Sigma-Aldrich) on SiO2 at 300°C was studied using in situ FTIR spectroscopy. The in situ FTIR studies were performed in a reactor equipped with an FTIR spectrometer that has been described previously.28 The FTIR experiments utilized high surface


Page 6 of 49

Page 7 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


area Si nanoparticles (>98%, US Research Nanomaterials) with an average diameter of 20 – 30 nm. The large surface area provided by the nanopowder sample improved the signal-to-noise ratio for infrared absorption. The Si nanoparticles were mechanically pressed into a tungsten grid support to facilitate the transmission FTIR measurements.29 The tungsten grid was 2 × 3 cm2, 50 µm thick, with 100 grid lines per inch. The tungsten grid could be resistively heated using a DC power supply (6268B, 20V/ 20A, Hewlett-Packard). The voltage output of the power supply was controlled by a PID temperature controller (Love Controls 16B, Dwyer Instruments). A type K thermocouple was attached to the bottom of the tungsten grid with epoxy (Ceramabond 571, Aremco) that attached and electrically isolated the thermocouple. To create a uniform SiO2 surface, the Si nanoparticles were annealed at 650°C for 1 hr under ~ 25 Torr of H2O vapor. Figure 1a shows the absolute FTIR spectra of the SiO2 coating on the Si nanoparticles before and after the H2O oxidation process. Distinctive absorption features for the unannealed sample are observed in four frequency regions: 835 – 1250 cm-1, 2080 – 2260 cm-1, 2855 – 2930 cm-1, and 3250 – 3750 cm-1. The strong and broad absorption bands between 835 – 1250 cm-1 are consistent with the TO/LO modes that arise from Si-O symmetric and asymmetric stretching vibrations.30 The absorption bands observed between 2080 – 2260 cm-1 are attributed to SiHx and SiOxH3-x species.31-33 Absorption bands between 2855 – 2930 cm-1 are attributed to the C-H stretching vibrations of adventitious hydrocarbons. The broad absorption feature observed between 3250 – 3750 cm-1 is consistent with isolated and hydrogen bonded Si-OH species. These results are in agreement with a partially oxidized hydrogen-terminated silicon surface. After thermal oxidation with H2O, evidence of silicon oxidation is observed in each distinctive absorption band. Between 835 – 1250 cm-1, large absorbance gains are observed that



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

are consistent with the addition of bulk Si-O-Si stretching vibrations. These results are consistent with previous studies that have demonstrated the formation of a thermal oxide layer on silicon surfaces exposed to water vapor at elevated temperatures.34-35 Furthermore, a distinct blue shift from 1045 to 1065 cm-1 and 1160 to 1245 cm-1 is observed for the TO/LO modes after thermal oxidation. This behavior has been observed previously and is related to the stoichiometry of the SiOx film. As more oxygen is incorporated into the film, a more electronegative environment leads to shorter Si-O bonds lengths and higher stretching frequencies.30 The disappearance of the vibrational features between 2080 – 2260 cm-1 is attributed to Si-H removal via H2 evolution. H2 desorption has been shown to occur on hydrogen-covered Si surfaces at 325-625°C.36-37 Thermal Si-H removal results in Si surface dangling bond sites that are susceptible to oxidation.32, 38-39 Absorption features between 2855 – 2930 cm-1 are also lost due to the volatilization of the carbon contaminants in the oxidizing environment. Lastly, there is a large absorbance gain observed at 3750 cm-1 that is attributed to the formation of isolated SiOH species. Previous studies have also observed the formation of isolated Si-OH species during oxidation of hydrogen-passivated silicon surfaces.32 This H2O thermal oxidation procedure consistently removed any carbon-related vibrational modes and resulted in a SiO2 coating on the Si nanoparticles with absorbances attributed only to the O–H hydroxyl stretching vibrations and bulk SiO2 vibrational features. Transmission electron microscopy (TEM) analysis was utilized to confirm the presence of the SiO2 coating on the Si nanoparticles. Figure 1b shows the TEM image of an annealed silicon nanoparticle with a SiO2 coating with a thickness of ~3.5 nm. The SiO2 coating is very uniform and conformal on the Si nanoparticle.


Page 8 of 49

Page 9 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


III. RESULTS AND DISCUSSION 1. X-Ray Reflectivity Measurements of SiO2 Etching. Figure 2 displays XRR measurements of the change in the SiO2 film thickness versus ALE cycles at 300°C for various static reactant pressures. The reaction sequence was defined by reactant exposures with a duration of 20 s and purges with a duration of 120 s. This reaction sequence is designated 20120-20-120. The reactant pressures for the TMA and HF exposures were the same. The initial SiO2 thermal oxide film on the Si(100) wafer had a thickness between 390 - 397 Å. Figure 2 shows that the SiO2 film thickness change versus number of ALE cycles is linear at each reactant pressure. The SiO2 etch rates were 0.027, 0.15, 0.20, and 0.31 Å /cycle for static reactant pressures of 0.1, 0.5, 1.0, and 4.0 Torr, respectively. Previous results that utilized lower reactant pressures of ~0.08 Torr in a viscous flow mode showed no SiO2 etching by TMA and HF at 300°C.18 The SiO2 etch rates observed at higher precursor pressures suggest a pressure-dependent reaction mechanism for SiO2 ALE. Control experiments were also performed using only HF exposures of 1.0 Torr. The results in Figure 2 indicate that HF exposures by themselves resulted in negligible etching of the SiO2 film at 300°C. These results are consistent with earlier results that concluded that HF alone cannot etch SiO2.7, 9 H2O vapor needs to be present together with HF for SiO2 etching. 2. Spectroscopic Ellipsometry Measurements of SiO2 Etching. Figure 3 displays SE measurements of the change in the SiO2 film thickness versus ALE cycles at 300°C for various static reactant pressures. The reaction sequence was again defined by reactant exposures with a duration of 20 s and purges with a duration of 120 s. The reactant pressures for the TMA and HF



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

exposures were the same. The initial SiO2 thermal oxide film on the Si(100) wafer again had a thickness between 390 - 397 Å. Figure 3 shows that the SiO2 film thickness change versus number of ALE cycles is linear at each reactant pressure. The etch rates were 0.07, 0.17, 0.18, and 0.27 Å /cycle for static reactant pressures of 0.1, 0.5, 1.0, and 4.0 Torr, respectively. The etch rates from the SE measurements are in good agreement with the etch rates from the XRR measurements. Control experiments were also performed using only HF exposures of 1.0 Torr. The results in Figure 3 again demonstrate that only HF exposures lead to negligible SiO2 etching. 3. FTIR Spectroscopy Measurements of SiO2 Etching. Figure 4 displays the absolute FTIR spectra from 700 – 1500 cm-1 for Si particles covered with a SiO2 coating during static TMA and HF exposures at 0.5, 1 and 4 Torr at 300°C. Each reactant exposure consisted of a static dose at the specified pressure for 30 s followed by a 240 s purge. The most prominent vibrational features are attributed to the transverse optical (TO) and longitudinal optical (LO) modes arising from asymmetric Si-O stretching vibrations of SiO4 tetrahedra at 1065 and 1230 cm-1, respectively.30 The less intense peak at 820 cm-1 is attributed to the corresponding TO and LO modes arising from symmetric Si-O stretching vibrations.30 The absorbance features from the Si-O stretching vibrations decrease progressively versus the number of ALE cycles. An absorbance feature also appears at 1275 cm-1 after the first TMA exposure that is attributed to a deformation mode of Si-CH3 surface species.40 This result is consistent with previous studies that have observed the formation of Si-CH3 species during reactions of TMA with SiO2 surfaces.41-43 The FTIR spectra in Figure 4 were recorded after the TMA exposures. The first 50 cycles consisted of static reactant exposures at 1 Torr and showed a progressive loss of the


Page 10 of 49

Page 11 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


absorbance for the Si-O stretching vibrations versus the number of ALE cycles. The nanoparticles were subsequently exposed to 30 cycles of static reactant exposures at 4 Torr and then 30 cycles of static reactant exposures at 0.5 Torr. The absorbance change per cycle of the Si-O stretching vibrations increased significantly at 4 Torr. This more rapid absorbance change indicates a higher SiO2 etch rate at the higher reactant pressure. The absorbance change per cycle then decreased during the subsequent 0.5 Torr static exposures. The absolute integrated absorbances of the asymmetric Si-O stretching vibrations between 900 – 1400 cm-1 versus ALE cycle number are plotted in Figure 5. At each pressure, the absorbance decreases linearly versus sequential TMA and HF exposures. ∆Abs/cycle values of -1.1, -1.9 and -3.6 are obtained for reactant pressures of 0.5, 1 and 4 Torr, respectively. These results are in agreement with the ex situ XRR and SE measurements that revealed larger SiO2 etch rates at higher reactant pressures. The etch rates and absorbance losses per cycle versus reactant pressure obtained by XRR, SE, and FTIR measurements are plotted in Figure 6. The relative etch rates determined by XRR and SE measurements at various reactant pressures are in good agreement with the absorbance loss per cycle determined using FTIR spectroscopy. All of the measurement techniques showed that the etch rate increases up to 1 Torr and then begins to level off at higher pressures. The leveling off of the etch rate at higher pressures suggests a diffusion-limited process similar to the oxidation of silicon substrates.34 The surface layers formed during either the TMA or HF reactions act as diffusion barriers. These surface layers progressively limit the transport of reactants as the thickness of the surface layer increases. 4. FTIR Studies of Surface Chemistry and Thin Film Changes. The FTIR spectra can examine the surface chemistry and thin film changes during SiO2 ALE. Figure 7 displays



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

the FTIR difference spectra between 500 – 1700 cm-1 for the first TMA and HF exposures on SiO2 at 300°C. Each reactant exposure consisted of a ~1 Torr static dose for 30 s followed by a purge for 240 s. Figure 7a shows the difference spectrum defined by the spectrum recorded after the first TMA exposure referenced to the spectrum of the initial SiO2 coating on the Si particles (TMA- SiO2). Negative absorbance features at 975 and 1250 cm-1 are observed that are consistent with the loss of Si-OH and Si-O species, respectively.30, 44 These absorbance losses are accompanied by absorbance gains at 670, 750, 845, 1212 and 1275 cm-1. The absorbance gains at 670 and 1275 cm-1 in Figure 7a are attributed to the addition of Si-CH3 species.41, 45 The absorbance gains at 750 and 1212 cm-1 are consistent with the addition of Al-CH3 species. The absorbance gain at 845 cm-1 is assigned to the formation of Al-O species.46-48 The resulting surface could contain AlO3-x(CH3)x species where x= 1 or 2 and SiO4-y(CH3)y species where y= 1, 2 or 3. These FITR measurements are in agreement with previous studies that have reported that TMA chemisorbs on silica surfaces primarily through reaction with hydroxyl groups and siloxane bridges.41, 43, 49 TMA reactions with hydroxyl groups result in the formation of siloxy-methylaluminum surface species and methane. During dissociation of TMA on siloxane bridges, methyl groups are transferred from aluminum to silicon. Figure 7b displays infrared difference spectrum after the subsequent HF exposure. The Si-O-Al and Al-O vibrational features that were added during the previous TMA exposure are lost after the subsequent HF exposure. These absorbance losses are accompanied by absorbance gains at 675 and 930 cm-1 that are consistent with Al-F and Si-F species, respectively.50-51 These results suggest that the Al-O species and Si-O-Al species formed during the TMA exposures are


Page 12 of 49

Page 13 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


converted to Al-F and Si-F species during HF exposures, respectively. The reaction of HF with Al2O3 to produce AlF3 has been observed in previous ALE studies.17-19 This reaction occurs as: Al2O3 + 6HF 2AlF3 + 3 H2O

(1)

Thermochemical calculations indicate that this reaction is spontaneous at 300°C with a Gibbs free energy value of ∆G = -41.1 kcal.52 Figure 8 shows the FTIR difference spectra between 500 – 1700 cm-1 for two consecutive TMA and HF exposures recorded during the 6th ALE cycle at 300°C. These difference spectra are referenced using the spectra after the previous reactant exposure and are displaced for clarity in presentation. Each reactant exposure consisted of a ~ 1 Torr static dose for 30 s followed by a purge for 240 s. Figure 8a displays the difference spectrum after the TMA exposure referenced to the spectrum collected after the previous HF exposure (TMA – HF). After TMA exposures, a broad absorbance loss is observed at 675 cm-1 that is consistent with the removal of Al-F stretching vibrations.51 Etching of AlF3 has been observed between 250 – 300°C using TMA.16, 18 The etching of AlF3 is consistent with the reaction: AlF3 + 2Al(CH3)3 → 3AlF(CH3)2

(2)

In this ligand-exchange reaction, TMA accepts fluorine from the AlF3 surface layer to form volatile AlF(CH3)2 species. AlF(CH3)2 is stable and has been shown to have a vapor pressure of 80 Torr at 100°C.53 Methyl groups from TMA are also transferred to the AlF3 surface layer to form additional volatile AlF(CH3)2 reaction products. This ligand-exchange reaction removes Al-F species from the surface. Negative absorbance features are also observed in Figure 8a at 1065 and 1220 cm-1 that are consistent with the loss of Si-O asymmetric stretching vibrations.30 These absorbance losses are accompanied by absorbance gains at 845 and 1000 cm-1 that are attributed to the addition of



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Page 14 of 49

Al-O and Al-O-Si stretching vibrations, respectively.46-48, 54-55 These absorbance changes are consistent with the reaction of TMA with SiO2 to form Al2O3 and aluminosilicate surface species. The reaction of TMA with SiO2 to form Al2O3 could occur according to: 4Al(CH3)3 + 3SiO2 → 2Al2O3 + 3Si(CH3)4

(3)

This reaction is predicted to be spontaneous with a Gibbs Free Energy change of ∆G = -235 kcal/mole at 300°C.52 The reaction between TMA and SiO2 has also been computed to be thermodynamically favorable using density functional theory methods.56 Infrared analysis, temperature-programmed desorption, and X-ray photoelectron spectroscopy have also been used to analyze the Si-methyl and Al-methyl surface species that result from the chemisorption of TMA on SiO2. Above 175°C, the aluminum alkyl surface complex has been shown to decompose, yielding more methyl groups bonded to silicon. At 325°C, Si-CH3 species were observed to react further to form volatile tetramethylsilane (Si(CH3)4) as suggested in Equation 3.41 Previous studies have also investigated the reaction of Al with SiO2.57-59 Aluminum films deposited on thermally grown SiO2 surfaces have been shown to reduce SiO2 and yield Al2O3 and Si according to: 4Al + 3SiO2 → 2Al2O3 + 3Si

(4)

This reaction is thermodynamically favorable with a Gibbs Free energy value of ∆G = -68 kcal/mole at 300°C.52 There are similarities between the reactions of TMA and Al with SiO2. During thermal ALE, TMA may act like metallic aluminum to produce Al2O3 as expected for thermite reactions.60 Figure 8b shows the absorbance changes after a subsequent HF exposure. An absorbance gain at 675 cm-1 is accompanied by absorbance losses at 845 and 1000 cm-1. These results are


Page 15 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


similar to those observed in Figure 7b and are consistent with HF exposures acting to fluorinate the Al2O3/aluminosilicate layer formed during TMA exposures. This fluorinated layer can then be removed during subsequent TMA exposures. TMA can also act to further displace silicon at the surface to form a new Al2O3/aluminosilicate layer that can then be etched during the subsequent ALE cycle. 5. FTIR Studies of Self-Limiting Behavior and Pressure Dependence. The selflimiting behavior of the TMA and HF reactions at a given reactant pressure was determined by monitoring the absorbance changes for the Si-O stretching vibrations between 1050 – 1225 cm-1 during TMA exposures and the Al-F stretching vibrations between 550 – 750 cm-1 during HF exposures. Figure 9 examines the self-limiting behavior of the SiO2 ALE reactions with TMA and HF at 300°C. Each data point was collected after a ~0.5 Torr static dose for 5 s. Figure 9a displays the absorbance change for the Si-O stretching vibration between 1050 – 1225 cm-1 after TMA exposures. Self-limiting behavior is observed after a TMA exposure of 15 Torr s. The surface Al2O3/aluminosilicate layer acts as a diffusion barrier that limits Al diffusion through the layer to reach unreacted SiO2. Figure 9b shows the absorbance change for the Al-F stretching vibrations between 550 – 750 cm-1 after HF exposures. Self-limiting behavior is observed after a HF exposure of 10 Torr s. The self-limiting behavior is consistent with the fluorination of the Al-O species that were added during previous TMA exposures. After all of the Al-O species are converted to Al-F species, no further reaction occurs. FTIR studies also observed the absorbance changes at different reactant pressures. Figure 10 displays the FTIR difference spectra between 450 – 1450 cm-1 for increasing TMA pressures at 300°C. The difference spectra are referenced to the previous HF exposure. This previous exposure consisted of a ~1 Torr static HF dose for 30 s. Figure 10a was obtained after a



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

self-passivating reaction with a TMA static exposure of 0.5 Torr. The resulting vibrational features are similar to those presented in Figure 8. The TMA exposure of 0.5 Torr shows the loss of Al-F species at 675 cm-1 and Si-O-Si species at 1220 cm-1. Absorbance gains are also observed at 845 and 1000 cm-1 that are consistent with the addition of Al-O and Si-O-Al species. Figure 10b and 10c show the resulting difference spectra when the TMA reactant pressure is increased to 1.0 Torr and 4.0 Torr, respectively. These difference spectra were recorded after self-passivating reactions with TMA static exposures of 1.0 and 4.0 Torr, respectively. Little change is observed in the Al-F vibrational region. This behavior suggests that Al-F species are removed from the surface before Al reacts with the underlying SiO2 layer. Further absorbance loss at 1220 cm-1, and absorbance gains at 845 and 1000 cm-1, are consistent with additional conversion of Si-O species to Al-O species at the higher TMA pressures. Figure 11 shows the FTIR difference spectra between 450 – 1450 cm-1 for increasing HF pressures at 300°C. The difference spectra are referenced to the previous TMA exposure. This previous exposure consisted of a ~1 Torr static TMA dose for 30 s. Figure 11a was obtained after a self-passivating reaction with a HF static exposure of 0.5 Torr. The HF exposure of 0.5 Torr causes absorbance losses at 845 and 1000 cm-1 and absorbance gains at 675 and 955 cm-1. These absorbance changes are consistent with the conversion of Al-O and Si-O-Al species to AlF and Si-F species, respectively. Figure 11b was collected after a self-passivating HF reaction when the HF pressure was increased to 1.0 Torr. Figure 11c was recorded after a self-passivating reaction when the HF pressure was increased further to 4.0 Torr. These spectra show additional Al-O and Si-O-Al to Al-F conversion in each absorbance region at the higher HF pressures. The absorbance changes observed after the HF exposure at 4.0 Torr are much less than the absorbance changes observed


Page 16 of 49

Page 17 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


after the HF exposures at 0.5 and 1.0 Torr. These results are consistent with a self-passivating fluorination reaction. The self-passivating Al2O3/aluminosilicate and fluoride surface layers are similar to the self-passivating oxide layers that form on metal61-62 or silicon34, 63-64 surfaces and protect the metal or silicon surfaces from further oxidation. These self-passivating reactions can be understood in terms of diffusion-limited growth. The surface layers act as diffusion barriers that limit the conversion and fluorination reactions. Diffusion-limited growth has been described earlier in the classic Deal-Grove treatment of silicon oxidation.34 In the Deal-Grove approach,34 the flux of reactants, F, is dependent on the concentration gradient across the SiO2 surface layer according to: F = D(C0-Cs)/x

(5)

where D is the diffusion coefficient, C0 is the concentration of reactants on the SiO2 surface, Cs is the concentration of reactants at the SiO2/Si interface and x is the thickness of the SiO2 surface layer. Solution of the diffusion equations reveals that the thickness of the SiO2 surface layer will increase parabolically versus time and display self-passivating behavior at longer times.34 The larger amounts of conversion and fluorination at higher pressures can also be understood in terms of diffusion-limited growth. The pressure dependence of the thickness is determined by C0, the concentration of reactants on the surface. C0 is defined by the reactant pressure, P, and the Henry’s Law constant, H, according to C0 = HP. Higher pressures lead to higher surface reactant concentrations. These higher concentrations produce larger concentration gradients across the surface layer and thicker surface layers. 6. XPS Studies of Silicon Oxidation State after TMA and HF Reactions with SiO2. The reaction of TMA with SiO2 given by Equation 3 reduces silicon because CH3 is much less



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

electronegative than oxygen. The CH3 group has an electronegativity of χ= 2.525.65 In contrast, oxygen has an electronegativity of χ= 3.5. The reaction of TMA with SiO2 will lead to more electron density on silicon in the SiO4-y(CH3)y suboxide products. The reduction of SiO2 and formation of Al2O3/aluminosilicates after TMA exposure were observed by ex situ XPS analysis. Figure 12 displays the Si 2p XPS spectra of (a) bare SiO2, (b) SiO2 after an 8 Torr TMA exposure for 5 minutes at 300°C, and (c) SiO2 after an 8 Torr TMA exposure for 5 minutes followed by a 4 Torr HF exposure for 10 minutes at 300°C. The higher reactant pressures and exposure times were chosen to ensure that changes in the surface species were detectable by XPS analysis. The Si 2p peaks in Figure 12 were deconvoluted into silicon oxidation states of 4+, 3+, 2+, 1+ and 0. The fits were constrained by the FWHM and binding energies of the individual Si oxidation states as described in Section II.1. The binding energies of the silicon oxidation states of 4+, 3+, 2+, 1+ and 0 are based on values for SiO2 and are shown by the dashed lines in Figure 12.25-27 Figure 12a depicts the SiO2 control sample with a broad peak at 104.3 eV. This peak position is consistent with literature values for SiO2 and silicon atoms in the Si4+ oxidation state.66 Figure 12b shows the results after the TMA exposure. The Si 2p peak shifts to lower binding energies centered around 102.5 eV that is composed of 3+, 2+ and 1+ oxidation states. The disappearance of the 4+ oxidation state indicates the TMA exposure reduces the Si4+ in SiO2 into Si suboxides. As suggested by the reaction in Equation 3, these silicon suboxides may be SiO3(CH3), SiO2(CH3)2 and SiO(CH3)3 surface species. Figure 12c displays the XPS results for the Si 2p peak after sequential TMA and HF exposures. The HF exposure shifts the Si 2p peak back to higher energies and can be


Page 18 of 49

Page 19 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


deconvoluted into silicon oxidation states of 4+, 3+ and 2+. These higher silicon oxidation states are consistent with HF reacting with SiCH3 surface species according to SiCH3 + HF → SiF + CH4. This reaction could form SiFO3, SiFO2(CH3) and SiFO(CH3)2 surface species with silicon oxidation states of 4+, 3+ and 2+, respectively. These higher oxidation states are expected because fluorine has a high electronegativity of χ= 4.0. In addition, Figure 12c also displays a higher energy peak at 108.8 eV. This higher energy peak can be attributed to a shake-up satellite peak.67 Shake-up satellite peaks can occur when the emitted photoelectron excites an ion. These shake-up peaks from Si 2p can be observed when Si is present in a siloxane polymer structure.67 The reduction of SiO2 and formation of Al2O3/aluminosilicates after TMA exposure were also monitored by XPS analysis of the Si 2p XPS spectra after a 4 Torr TMA exposure on SiO2 for 20 s at 300°C. The Si 2p XPS spectra displayed a shift from the 4+ oxidation state for SiO2 to a distribution of 4+, 3+, 2+ and 1+ oxidation states after the TMA exposure. The peak of the spectrum occurred at the 3+ oxidation state compared with a peak at the 2+ oxidation state for the TMA exposure at 8 Torr shown in Figure 12b. The Si 2p XPS spectrum after a 4 Torr TMA exposure on SiO2 for 20 s followed by a 4 Torr HF exposure for 20 s at 300°C displayed a shift back to higher oxidation states similar to the results shown in Figure 12c. The Al 2p XPS peak was also consistent with the formation of Al2O3 at the surface after TMA exposure. Figures 13a and 13b show the Al 2p XPS spectra after the TMA exposure and after sequential TMA and HF exposures, respectively. These XPS spectra were obtained from the same samples as shown in Figure 12. Figure 13a displays a peak at 74.3 eV after the TMA exposure that corresponds to the formation of Al2O3 or aluminosilicates.66, 68 There is also a smaller shoulder on this peak at ~72 eV that corresponds to more metallic Al. This peak may result from some TMA decomposition.69



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure 13b shows that the Al 2p XPS peak shifts to higher binding energies after sequential TMA and HF exposures. The broader peak at ~76 eV is consistent with the fluorination of Al2O3.68 This peak may be composed of AlF3 at 77 eV and oxyaluminum fluoride species at lower energies.66, 68 Oxygen is less electronegative and will not shift the Al 2p binding energy as much as fluorine. 7. Mechanism for SiO2 ALE and Future Possibilities. A simplified schematic for the proposed SiO2 ALE chemistry during the sequential TMA and HF exposures on SiO2 is shown in Figure 14. After the first TMA and HF exposures, subsequent TMA exposures react with the AlF3 species on the SiO2 substrate to form volatile AlF(CH3)2 etch products. After removal of the AlF3 species, TMA reacts with the underlying SiO2 to form Al2O3 and aluminosilicates with the possible release of Si(CH3)4 or Si(CH3)xF4-x, as volatile reaction products. Mass spectrometry studies are required to confirm the volatile reaction products. In addition, suboxides such as AlO3-x(CH3)x where x= 1 or 2 and SiO4-y(CH3)y where y= 1, 2 or 3 may be present in the surface layer. The subsequent HF exposures converts the Al2O3 and aluminosilicates to AlF3 and releases H2O as a volatile reaction product. The AlO3-x(CH3)x and SiO4-y(CH3)y suboxides may also lead to the formation of CH4 during the HF exposure. Moreover, silicon species such as SiFO3, SiFO2(CH3) and SiFO(CH3)2 may be produced in the near surface layer. The shake-up satellite XPS peak from Si 2p also suggests that the Si may be in polymer structures similar to siloxane. The basics of this mechanism are similar to the recently observed “conversion-etch” mechanism for ZnO ALE.23 For ZnO ALE, TMA exposures converted ZnO to Al2O3. HF exposures then fluorinated the Al2O3 to AlF3. Subsequent TMA exposures removed the AlF3


Page 20 of 49

Page 21 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


and converted more ZnO to Al2O3.23 ZnO ALE by the “conversion-etch” mechanism occurred at lower reactant pressures of 60-80 mTorr. In contrast to the results for SiO2 ALE, higher reactant pressures > 0.1 Torr were not required for ZnO ALE. The lower pressures required for ZnO ALE suggests that “conversion-etch” may be more favorable for ZnO ALE than SiO2 ALE. However, thermochemical calculations do not support this viewpoint. The free energy change for the reaction 3ZnO + 2Al(CH3)3 → Al2O3 + 3Zn(CH3)2 is ∆G = -168 kcal at 300°C.52 In comparison, the free energy change for the reaction 1.5SiO2 + 2Al(CH3)3 → Al2O3 + 1.5Si(CH3)4 is ∆G = -197 kcal at 300°C.52 There must be other kinetic factors that favor ZnO ALE by the “conversion-etch” mechanism. The “conversion-etch” mechanism may be general and may be realized when the metal oxide formed by the conversion reaction is more stable than the initial metal oxide. Guidelines for determining the relative stability of metal oxides can be obtained from thermochemical calculations or Ellingham diagrams.70 Some materials may not have viable pathways for thermal ALE because the ligand-exchange reactions may not produce stable and volatile reaction products.20 By converting the initial material to a different material, the “conversion-etch” mechanism may enlarge the number of materials that can be etched using thermal ALE. For example, TMA is anticipated to convert other metal oxides, such as SnO2 and In2O3, to Al2O3.23 Thermal Al2O3 ALE has been demonstrated using HF and either Sn(acac)2 or TMA as the reactants.17-19 Likewise, TMA is expected to convert various nitrides, such as Si3N4, to AlN.23 Thermal AlN ALE has been demonstrated using HF and Sn(acac)2 as the reactants.21 Thermal SiO2 ALE also suggests a possible procedure for thermal Si ALE. The silicon surface can first be oxidized to form a self-passivating oxide layer on the silicon surface.63-64 The resulting SiO2 layer can then be etched by the “conversion-etch” mechanism using



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

sequential TMA and HF exposures. By repeating the silicon oxidation and sequential TMA and HF exposures, silicon should be etched with atomic layer precision. Similar procedures may also be defined for the thermal ALE of metals using sequential oxidation, fluorination and ligandexchange reactions.

IV. CONCLUSIONS Sequential reactions of trimethylaluminum (TMA) and hydrogen fluoride (HF) were employed for the thermal atomic layer etching (ALE) of SiO2 at 300°C. The etch rates during SiO2 ALE were dependent on reactant pressures. SiO2 etch rates of 0.027, 0.15, 0.20, and 0.31 Å/cycle were observed at static reactant pressures of 0.1, 0.5, 1.0 and 4.0 Torr, respectively, by ex situ x-ray reflectivity measurements. A similar increase in SiO2 etch rate with reactant pressure was confirmed by ex situ spectroscopic ellipsometry measurements. The pressure dependence of the etch rate for SiO2 ALE was also observed by in situ FTIR studies. In addition, the FTIR measurements monitored the pressure-dependent conversion of the SiO2 surface to an Al2O3/aluminosilicate layer. Ex situ XPS measurements also monitored the reduction of the silicon oxidation state after the TMA reaction. The Al2O3/aluminosilicate intermediate is attributed to the conversion of SiO2 to Al2O3 and reduced silicon species by the TMA exposure according to reactions like 3SiO2 + 4Al(CH3)3 → 2Al2O3 + 3Si(CH3)4. The HF exposure can then fluorinate Al2O3 to AlF3 and form silicon fluoride species after this conversion reaction. Subsequently, the next TMA exposure can remove AlF3 by a ligand-exchange transmetalation reaction and then convert additional SiO2 to Al2O3 and reduced silicon species. The sequential TMA and HF reactions lead to SiO2 ALE


Page 22 of 49

Page 23 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


because the individual TMA and HF reactions display self-limiting behavior at the various reactant pressures. The pressure-dependent conversion reaction of the SiO2 surface to an Al2O3/aluminosilicate layer by TMA is critical for thermal SiO2 ALE. The “conversion-etch” mechanism for SiO2 ALE is observed at reactant pressures > 0.1 Torr. Other conversion reactions will be able to transform initial materials that cannot be etched to different materials that may have possible etching pathways. The “conversion-etch” mechanism will increase the number of materials that can be etched by thermal ALE methods. Oxidation and conversion reactions may also lead to procedures for thermal Si ALE.

ACKNOWLEDGEMENTS This research was funded by the National Science Foundation by grants CHE-1306131 and CHE-1609554 and the Defense Advanced Projects Agency (DARPA) by grant W911NF-131-0041. The authors thank Huaxing Sun for the XPS measurements. The authors also thank Joel Clancey for assistance with the TEM analysis of the silicon nanoparticles.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

REFERENCES 1.

Nojiri, K. Dry Etching Technology for Semiconductors. Springer: Heidelberg, 2012.

2.

Carver, C. T.; Plombon, J. J.; Romero, P. E.; Suri, S.; Tronic, T. A.; Turkot, R. B.

Atomic Layer Etching: An Industry Perspective. ECS J. Solid State Sci. Technol. 2015, 4, N5005-N5009. 3.

Blumberg, A. A.; Stavrinou, S. C. Tabulated Functions for Heterogeneous Reaction

Rates - The Attack of Vitreous Silica by Hydrofluoric Acid. J. Phys. Chem. 1960, 64, 14381442. 4.

Monk, D. J.; Soane, D. S.; Howe, R. T. A Review of the Chemical Reaction

Mechanisma and Kinetics for Hydrofluoric Acid Etching of Silicon Dioxide for Surface Micromachining Applications. Thin Solid Films 1993, 232, 1-12. 5.

Spierings, G. A. C. M. Wet Chemical Etching of SIlicate-Glasses in Hydrofluoric Acid

Based Solutions. J. Mater. Sci. 1993, 28, 6261-6273. 6.

Knotter, D. M. Etching Mechanism of Vitreous Silicon Dioxide in HF-Based Solutions.

J. Am. Chem. Soc. 2000, 122, 4345-4351. 7.

Helms, C. R.; Deal, B. E. Mechanisms of the HF/H2O Vapor Phase Etching of SiO2. J.

Vac. Sci. Technol. A 1992, 10, 806-811. 8.

Vanderheide, P. A. M.; Hofman, M. J. B.; Ronde, H. J. Etching of Thin SiO2 Layers

Using Wet HF Gas. J. Vac. Sci. Technol. A 1989, 7, 1719-1723. 9.

Kang, J. K.; Musgrave, C. B. The Mechanism of HF/H2O Chemical Etching of SiO2. J.

Chem. Phys. 2002, 116, 275-280. 10.

Mogab, C. J.; Adams, A. C.; Flamm, D. L. Plasma Etching of Si and SiO2 - Effect of

Oxygen Additions to CF4 Plasmas. J. Appl. Phys. 1978, 49, 3796-3803.


Page 24 of 49

Page 25 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


11.

Rueger, N. R.; Beulens, J. J.; Schaepkens, M.; Doemling, M. F.; Mirza, J. M.; Standaert,

T.; Oehrlein, G. S. Role of Steady State Fluorocarbon Films in the Etching of Silicon Dioxide using CHF3 in an Inductively Coupled Plasma Reactor. J. Vac. Sci. Technol. A 1997, 15, 18811889. 12.

Zhang, D.; Kushner, M. J. Investigations of Surface Reactions during C2F6 Plasma

Etching of SiO2 with Equipment and Feature Scale Models. J. Vac. Sci. Technol. A 2001, 19, 524-538. 13.

Metzler, D.; Bruce, R. L.; Engelmann, S.; Joseph, E. A.; Oehrlein, G. S. Fluorocarbon

Assisted Atomic Layer Etching of SiO2 using Cyclic Ar/C4F8 Plasma. J. Vac. Sci. Technol. A 2014, 32, 020603. 14.

Oehrlein, G. S.; Metzler, D.; Li, C. Atomic Layer Etching at the Tipping Point: An

Overview. ECS J. Solid State Sci. Technol. 2015, 4, N5041-N5053. 15.

Lee, Y.; DuMont, J. W.; George, S. M. Atomic Layer Etching of HfO2 Using Sequential,

Self-Limiting Thermal Reactions with Sn(acac)2 and HF. ESC J. Solid State Sci. Technol. 2015, 4, N5013-N5022. 16.

Lee, Y.; DuMont, J. W.; George, S. M. Atomic Layer Etching of AlF3 Using Sequential,

Self-Limiting Thermal Reactions with Sn(acac)2 and Hydrogen Fluoride. J. Phys. Chem. C 2015, 119, 25385-25393. 17.

Lee, Y.; DuMont, J. W.; George, S. M. Mechanism of Thermal Al2O3 Atomic Layer

Etching Using Sequential Reactions with Sn(acac)2 and HF. Chem. Mater. 2015, 27, 3648-3657. 18.

Lee, Y.; DuMont, J. W.; George, S. M. Trimethylaluminum as the Metal Precursor for

the Atomic Layer Etching of Al2O3 Using Sequential, Self-Limiting Thermal Reactions. Chem. Mater. 2016, 28, 2994-3003.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

19.

Lee, Y.; George, S. M. Atomic Layer Etching of Al2O3 Using Sequential, Self-Limiting

Thermal Reactions with Sn(acac)2 and HF. ACS Nano 2015, 9, 2061-2070. 20.

Lee, Y.; Huffman, C.; George, S. M. Selectivity in Thermal Atomic Layer Etching Using

Sequential, Self-Limiting Fluorination and Ligand-Exchange Reactions. Chem. Mater. 2016, 28, 7657-7665. 21.

Johnson, N. R.; Sun, H.; Sharma, K.; George, S. M. Thermal Atomic Layer Etching of

Crystalline Aluminum Nitride Using Sequential, Self-Limiting Hydrogen Fluoride and Sn(acac)2 Reactions and Enhancement by H2 and Ar Plasmas. J. Vac. Sci. Technol. A 2016, 34, 050603. 22.

George, S. M.; Lee, Y. Prospects for Thermal Atomic Layer Etching Using Sequential,

Self-Limiting Fluorination and Ligand-Exchange Reactions. ACS Nano 2016, 10, 4889-4894. 23.

Zywotko, D. R.; George, S. M. Thermal Atomic Layer Etching of ZnO by a

“Conversion-Etch” Mechanism Using Sequential Exposures of Hydrogen Fluoride and Trimethylaluminum. Chem. Mater. 2017, 29, 1183-1191. 24.

Elam, J. W.; Groner, M. D.; George, S. M. Viscous Flow Reactor with Quartz Crystal

Microbalance for Thin Film Growth by Atomic Layer Deposition. Rev. Sci. Instrum. 2002, 73, 2981-2987. 25.

Bell, F. G.; Ley, L. Photoemission Study of SiOx (0 < x < 2) Alloys. Phys. Rev. B 1988,

37, 8383-8393. 26.

Saito, Y. Adsorption of Anhydrous Hydrogen Fluoride onto Silicon and Native Oxide by

XPS. Surf. Sci. Spectra 1999, 6, 228-236. 27.

Kim, S.; Kim, M. C.; Choi, S. H.; Kim, K. J.; Hwang, H. N.; Hwang, C. C. Size

Dependence of Si 2p Core-level Shift at Si Nanocrystal/SiO2 Interfaces. Appl. Phys. Lett. 2007, 91, 103113.


Page 26 of 49

Page 27 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


28.

DuMont, J. W.; George, S. M. Pyrolysis of Alucone Molecular Layer Deposition Films

Studied Using In Situ Transmission Fourier Transform Infrared Spectroscopy. J. Phys. Chem. C 2015, 119, 14603-14612. 29.

Ferguson, J. D.; Weimer, A. W.; George, S. M. Atomic Layer Deposition of Ultrathin

and Conformal Al2O3 films on BN Particles. Thin Solid Films 2000, 371, 95-104. 30.

Queeney, K. T.; Weldon, M. K.; Chang, J. P.; Chabal, Y. J.; Gurevich, A. B.; Sapjeta, J.;

Opila, R. L. Infrared Spectroscopic Analysis of the Si/SiO2 Interface Structure of Thermally Oxidized Silicon. J Appl. Phys. 2000, 87, 1322-1330. 31.

Devine, R. A. B. Structural Nature of the Si/SiO2 Interface Through Infrared

Spectroscopy. Appl. Phys. Lett. 1996, 68, 3108-3110. 32.

Mawhinney, D. B.; Glass, J. A.; Yates, J. T. FTIR Study of the Oxidation of Porous

Silicon. J. Phys. Chem. B 1997, 101, 1202-1206. 33.

Ogata, Y.; Niki, H.; Sakka, T.; Iwasaki, M. Oxidation of Porous Silicon Unver Water

Vapor Environment. J. Electrochem. Soc. 1995, 142, 1595-1601. 34.

Deal, B. E.; Grove, A. S. General Relationship for Thermal Oxidation of Silicon. J. Appl.

Phys. 1965, 36, 3770-3778. 35.

Bhat, V. K.; Pattabiraman, M.; Bhat, K. N.; Subrahmanyam, A. The Growth of Ultrathin

Oxides of Silicon by Low Temperature Wet Oxidation Technique. Mater. Res. Bull. 1999, 34, 1797-1803. 36.

Gupta, P.; Colvin, V. L.; George, S. M. Hydrogen Desorption Kinetics from

Monohydride and Dihydride Species on Silicon Surfaces. Phys. Rev. B 1988, 37, 8234-8243.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

37.

Koehler, B. G.; Mak, C. H.; Arthur, D. A.; Coon, P. A.; George, S. M. Desorption

Kinetics of Hydrogen and Deuterium from Si(111)7x7 Studied Using Laser Induced Thermal Desorption. J. Chem. Phys. 1988, 89, 1709-1718. 38.

Gupta, P.; Mak, C. H.; Coon, P. A.; George, S. M. Oxidation Kinetics of Si(111)7x7 in

the Submonolayer Regime. Phys. Rev. B 1989, 40, 7739-7749. 39.

Koehler, B. G.; Mak, C. H.; George, S. M. Decomposition of H2O on Si(111)7x7

Studied Using Laser Induced Thermal Desorption. Surf. Sci. 1989, 221, 565-589. 40.

Bellamy, L. J. The Infra-red Spectra of Complex Molecules. Chapman and Hall: London,

1975. 41.

Bartram, M. E.; Michalske, T. A.; Rogers, J. W. A Reexamination of the Chemisorption

of Trimethylaluminum on Silica. J. Phys. Chem. 1991, 95, 4453-4463. 42.

Puurunen, R. L.; Root, A.; Haukka, S.; Iiskola, E. I.; Lindblad, M.; Krause, A. O. I. IR

and NMR Study of the Chemisorption of Ammonia on Trimethylaluminum-Modified Silica. J. Phys. Chem. B 2000, 104, 6599-6609. 43.

Yates, D. J. C.; Dembinsk, G. W.; Kroll, W. R.; Elliott, J. J. Infrared Studies of

Reactions between Silica and Trimethylaluminum. J. Phys. Chem. 1969, 73, 911-921. 44.

Morrow, B. A.; McFarlan, A. J. Surface Vibrational Modes of Silanol Groups on Silica.

J. Phys. Chem. 1992, 96, 1395-1400. 45.

Grill, A.; Neumayer, D. A. Structure of Low Dielectric Constant to Extreme Low

Dielectric Constant SiCOH Films: Fourier Transform Infrared Spectroscopy Characterization. J. Appl. Phys. 2003, 94, 6697-6707.


Page 28 of 49

Page 29 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


46.

Frank, M. M.; Chabal, Y. J.; Green, M. L.; Delabie, A.; Brijs, B.; Wilk, G. D.; Ho, M. Y.;

da Rosa, E. B. O.; Baumvol, I. J. R.; Stedile, F. C. Enhanced Initial Growth of Atomic Layer Deposited Metal Oxides on Hydrogen-Terminated Silicon. Appl. Phys. Lett. 2003, 83, 740-742. 47.

Frank, M. M.; Chabal, Y. J.; Wilk, G. D. Nucleation and Interface Formation

Mechanisms in Atomic Layer Deposition of Gate Oxides. Appl. Phys. Lett. 2003, 82, 4758-4760. 48.

Mackenzie, K. J. Infrared Frequency Calculations for Ideal Mullite (3Al2O3.2SiO2). J.

Am. Ceram. Soc. 1972, 55, 68-71. 49.

Kinney, J. B.; Staley, R. H. Reactions of Titanium Tetrachloride and Trimethylaluminum

at Silica Surfaces Studies by Using Infrared Photo-acoustic Spectroscopy. J. Phys. Chem. 1983, 87, 3735-3740. 50.

Barabash, R. M.; Zaitsev, V. N.; Kovalchuk, T. V.; Sfihi, H.; Fraissard, J. Low-

Temperature Fluorination of Silica by a Nonaqueous Solution of NH4F. J. Phys. Chem. A 2003, 107, 4497-4505. 51.

Gross, U.; Ruediger, S.; Kemnitz, E.; Brzezinka, K.-W.; Mukhopadhyay, S.; Bailey, C.;

Wander, A.; Harrison, N. Vibrational Analysis Study of Aluminum Trifluoride Phases. J. Phys. Chem. A 2007, 111, 5813-5819. 52.

HSC Chemistry. HSC Chemistry 5.1, Outokumpu Research Oy: Pori, Finland.

53.

Weidlein, J.; Krieg, V. Vibrational Spectra of Dimethyl and Diethyl Aluminum Fluoride.

J. Organomet. Chem. 1968, 11, 9-16. 54.

Morozova, N. B.; Zherikova, K. V.; Baidina, I. A.; Sysoev, S. V.; Semyannikov, P. P.;

Yakovkina, L. V.; Smirnova, T. P.; Gelfond, N. V.; Igumenov, I. K.; Carta, G.; Rossetto, G. Volatile Hafnium(IV) Compounds with Beta-Diketonate and Cyclopentadienyl Derivatives. J. Phys. Chem. Solids 2008, 69, 673-679.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

55.

Mozgawa, W.; Sitarz, M.; Rokita, M. Spectroscopic Studies of Different Aluminosilicate

Structures. J. Mol. Struct. 1999, 511, 251-257. 56.

Lamagna, L.; Wiemer, C.; Perego, M.; Spiga, S.; Rodriguez, J.; Coll, D. S.; Grillo, M. E.;

Klejna, S.; Elliott, S. D. Mechanisms for Substrate-Enhanced Growth during the Early Stages of Atomic Layer Deposition of Alumina onto Silicon Nitride Surfaces. Chem. Mater. 2012, 24, 1080-1090. 57.

Bauer, R. S.; Bachrach, R. Z.; Brillson, L. J. Au and Al Interface Reactions with SiO2.

Appl. Phys. Lett. 1980, 37, 1006-1008. 58.

Hecht, M. H.; Vasquez, R. P.; Grunthaner, F. J.; Zamani, N.; Maserjian, J. A Novel X-

ray Photoelectron Spectroscopy Study of the Al/SiO2 Interface. J. Appl. Phys. 1985, 57, 52565262. 59.

Tsukada, M.; Ohfuji, S. I. In Situ X-Ray Photoemission Spectroscopic Studies of

Al/SiO2 Interface Formation. J. Vac. Sci. Technol. A 1994, 12, 1-6. 60.

Wang, L. L.; Munir, Z. A.; Maximov, Y. M. Thermite Reactions - Their Utilization in

the Synthesis and Processing of Materials. J. Mater. Sci. 1993, 28, 3693-3708. 61.

Graham, M. J.; Cohen, M. On the Mechanism of Low Temperature Oxidation (23-450

C) of Polycrystalline Nickel. J. Electrochem. Soc. 1972, 119, 879-882. 62.

Song, S.; Placido, F. Investigation on Initial Oxidation Kinetics of Al, Ni, and Hf Metal

Film Surfaces. Chin. Opt. Lett. 2010, 8, Supplement, 87-90. 63.

Derrien, J.; Commandre, M. SiO2 Ultra Thin Film Growth Kinetics as Investigated by

Surface Techniques. Surf. Sci. 1982, 118, 32-46. 64.

Fehlner, F. P. Formation of Ultrathin Oxide Films on Silicon. J. Electrochem. Soc. 1972,

119, 1723-1727.


Page 30 of 49

Page 31 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


65.

Matsunaga, N.; Rogers, D. W.; Zavitsas, A. A. Pauling's Electronegativity Equation and

a New Corollary Accurately Predict Bond Dissociation Enthalpies and Enhance Current Understanding of the Nature of the Chemical Bond. J. Org. Chem. 2003, 68, 3158-3172. 66.

NIST X-ray Photoelectron Spectroscopy Database, Version 4.1 (National Institute of

Standards and Technology, Gaithersburg, Maryland, 2012); http://srdata.nist.gov/xps/. 67.

Gardella, J. A.; Ferguson, S. A.; Chin, R. L. π* ← π Shakeup Satellites for the Analysis

of Structure and Bonding in Aromatic Polymers by X-ray Photoelectron Spectroscopy. Appl. Spectrosc. 1986, 40, 224-232. 68.

Bose, O.; Kemnitz, E.; Lippitz, A.; Unger, W. E. S. C 1s and Au 4f7/2 Referenced XPS

Binding Energy Data Obtained with Different Aluminium Oxides, -Hydroxides and -Fluorides. Fresenius J. Anal. Chem. 1997, 358, 175-179. 69.

Mayer, T. M.; Rogers, J. W.; Michalske, T. A. Mechanism of Nucleation and Atomic

Layer Growth of AlN on Si. Chem. Mater. 1991, 3, 641-646. 70.

Ellingham, H. J. T. Reducibility of Oxides and Sulfides in Metallurgical Processes. J.

Soc. Chem. Ind. London 1944, 63, 125-133.



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

Figure Captions Figure 1

(a) Absolute FTIR spectra the silicon particles with a SiO2 coating before and after oxidation with ~25 Torr H2O at 650°C for 1 hr. (b) TEM image of a silicon nanoparticle with a ~3.5 nm thick SiO2 coating after the H2O thermal oxidation.

Figure 2

SiO2 film thickness change versus number of SiO2 ALE cycles determined by Xray reflectivity measurements at TMA/HF pressures of 0.1, 0.5, 1.0 and 4.0 Torr. HF only experiments were also performed using an HF pressure of 1.0 Torr.

Figure 3

SiO2 film thickness change versus number of SiO2 ALE cycles determined by spectroscopic ellipsometry measurements at TMA/HF pressures of 0.1, 0.5, 1.0 and 4.0 Torr. HF only experiments were also performed using an HF pressure of 1.0 Torr.

Figure 4

Absolute FTIR spectra showing the absorbance loss from Si-O stretching vibrations between 700 – 1500 cm-1 versus SiO2 ALE cycles at TMA/HF pressures of 0.5, 1.0 and 4.0 Torr at 300°C.

Figure 5

Absolute integrated absorbance of the Si-O stretching vibrations between 900 – 1400 cm-1 versus ALE cycle number at TMA/HF pressures of 0.5, 1.0 and 4.0 Torr. The integrated absorbance change per cycle was -1.1, -1.9 and -3.6 at pressures of 0.5, 1.0 and 4.0 Torr, respectively.

Figure 6

Etch rates from the XRR measurements in Figure 2 and the SE measurements in Figure 3 compared with the absorbance loss versus cycle from Figure 5 versus reactant pressure.


Page 32 of 49

Page 33 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


Figure 7

Difference infrared absorbance spectra from 500 – 1700 cm-1 during the first TMA and HF exposures on SiO2 at 300°C. (a) Spectrum after the first TMA exposure referenced to the spectrum for the initial SiO2 layer on the silicon particles. (b) Spectrum after the first HF exposure referenced to the spectrum after the first TMA exposure.

Figure 8

Difference infrared absorbance spectra from 500 – 1700 cm-1 during the 6th ALE cycle at 300°C. (a) Spectrum after the first 6th TMA exposure referenced to the spectrum after the 5th HF exposure. (b) Spectrum after the first 6th HF exposure referenced to the spectrum after the 6th HF exposure.

Figure 9

(a) Normalized integrated absorbance of the Si-O stretching vibration between 1050 – 1225 cm-1 during the TMA exposure. (b) Normalized integrated absorbance of the Al-F stretching vibration between 550 – 750 cm-1 during the HF exposure.

Figure 10

Difference infrared absorption spectra between 450 – 1450 cm-1 for increasing TMA pressures of (a) 0.5 (b) 1.0 and (c) 4.0 Torr at 300°C. Difference spectra are referenced to the spectrum after the previous HF exposure at 1 Torr for 30 s.

Figure 11

Difference infrared absorption spectra between 450 – 1450 cm-1 for increasing HF pressures of (a) 0.5 (b) 1.0 and (c) 4.0 Torr at 300°C. Difference spectra are referenced to the spectrum after the previous TMA exposure at 1 Torr for 30 s.

Figure 12

XPS results for the Si 2p peak from (a) initial SiO2 film; (b) SiO2 film after TMA exposure; and (c) SiO2 film after TMA and HF exposures. The dashed lines



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60

correspond to the 4+, 3+, 2+, 1+ and 0 oxidation states for the Si 2p peak. The Si 2p peaks are deconvoluted using previously measured FWHM values. Figure 13

XPS results for the Al 2p peak from (a) SiO2 film after TMA exposure and (b) SiO2 film after TMA and HF exposures. These XPS spectra were obtained from the same samples as shown in Figure 12. The dashed line corresponds to the binding energy for Al in Al2O3.

Figure 14

Schematic of the proposed reaction mechanism for SiO2 ALE showing the sequential TMA and HF reactions.


Page 34 of 49

Page 35 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60


TOC Graphic:



1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58

Page 36 of 49

(b)

ACS Paragon Plus Environment

Figure 1

Page 37 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58



Figure 2


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58


Page 38 of 49

Figure 3

Page 39 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58



Figure 4


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58


Page 40 of 49

Figure 5

Page 41 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58



Figure 6


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58


Page 42 of 49

Figure 7

Page 43 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58



Figure 8


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58


Page 44 of 49

Figure 9

Page 45 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58



Figure 10


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58


Page 46 of 49

Figure 11

Page 47 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58



Figure 12


1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58


Page 48 of 49

Figure 13

Page 49 of 49

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58



Figure 14

Thermal Atomic Layer Etching of SiO2 by a âConversion-Etch

Recommend Documents

Thermal Atomic Layer Etching of SiO2 by a âConversion-Etch